Scott Sumner  

Kevin Drum's Third Law is wrong

John Thacker on Democrats Subs... The Incredible Vanishing Minar...

Tyler Cowen recently listed three laws, and Kevin Drum responded with variations on Tyler's Laws. Here was Tyler's third law:

3. Cowen's Third Law: All propositions about real interest rates are wrong.
And Kevin's version:
3. Drum's Third Law: Really? Isn't there a correlation between real interest rates and future inflationary expectations? In general, don't low real interest rates make capital investment more likely by lowering hurdle rates? Or am I just being naive here?
Kevin is reasoning from a price change. When real interest rates fall, investment usually declines as well. Never reason from a price change, always reason from the thing that causes the price change. In the case of real interest rates it's usually the business cycle. The most common cause of low real interest rates is recessions, and recessions are also associated with low levels of investment.

Can this claim be saved by talking about changes in interest rates "other things equal"? No, because "other things equal" interest rates never change.

Kevin Drum should not feel bad, however, as Nobel Prize winner Robert Shiller recently made the exact same mistake.

Is Tyler Cowen's Third Law also wrong? In a sense I've provided support for his third law, by finding a flaw in Drum's version. And yet I'm also making claims about real interest rates. Indeed in a sense Tyler is as well. Or is he making claims about claims? I'll let the philosophers sort this one out.

Comments and Sharing

CATEGORIES: Macroeconomics

COMMENTS (7 to date)
Michael Byrnes writes:

And even Keynes made the mistake, right?

Andrew_FL writes:

He is in fact making a claim about propositions-specifically the class of propositions consisting of those which concern real interest rates. That's not really a question of philosophy though, so much as it is grammar.

ThomasH writes:
No, because "other things equal" interest rates never change.

Does this mean that real interest rates would not (given some monetary rule) change in response to a change in inflationary expectations or a claim that there can be no exogenous change in inflationary expectations?

Daniel Kuehn writes:

Michael Byrnes - I'm curious what Scott thinks of that but I think an important distinction with Keynes's comment is that he's talking about a "lower" interest rate, not a "low" interest rate. It's a relative claim, he's not talking about the price level per se. We can accuse him of being vague of course - lower than what? - but presumably he's got something in mind about departures from a natural rate or a full employment rate or something like that. Vague, but not quite the same as Drum in my opinion.

I actually think "reasoning from a price change" tends to be over-diagnosed among economists at least. Often people are thinking of some reference point and either the reference point is implicit (as in Keynes), or they're just being sloppy with language but not really reasoning from a price change.

Jose Romeu Robazzi writes:

Prof. Sumner
This is a very insteresting point. I have always been uneasy with the idea that "low rates induce investment because the hurdle rate is lower". This link may exist, but companies don't really undertake new projects just because it is cheaper to finance them. This may be true in purely financial projects, but not on Main Street. On the contrary, what Main Street picks up is higher demand for their products, and this is why I think the hypothesis under your "hot potatoes" effect may well be the best explanation why monetary policy works.

Scott Sumner writes:

Michael, Keynes is clearly wrong, but I don't know that the mistake is reasoning from a price change.

Thomas, In general, a change in inflation expectations will generally affect nominal rates, not real rates. There may be cases where it affects real rates, but in those cases something else needs to be happening.

Daniel, When people clearly have a causal factor in mind I give them a pass. I only point to reasoning from a price change when the reasoning process leads them to the wrong conclusion, as in this case.

Thanks Jose.

maynardGkeynes writes:

Speaking of Shiller, I believe that his research showed that there has been very little correlation between mortgage interest rates and home price movements historically. This suggests to me that he actually agrees with you on not reasoning from a price change, at least with regard to interest rates.

Comments for this entry have been closed
Return to top